A New Kernel Based on Weighted Cross-Correlation Coefficient for SVMs and Its Application on Prediction of T-cell Epitopes

摘要

T-cell epitopes play vital roles in immune response. Its recognition by T-cell receptors is a precondition for the activation of T-cell clone. This recognition is antigen-specific. Therefore, identifying the pattern of a MHC restricted T-cell epitopes is of great importance for immunotherapy and vaccine design. In this paper, we designed a new kernel based on weighted crosscorrelation coefficients for support vector machine and applied it to the direct prediction of T-cell epitopes. The experiment was carried on an MHC type I restricted T-cell clone LAU203-1.5. The results showed that this approach is efficient and promising.